Yann LeCun
Chief AI Scientist · Turing Award Winner; intellectual architect of Meta’s open-weights model strategy (Llama 1, 2, 3) and Joint Embedding Predictive Architecture (JEPA).
Stance on AGI
"Argues current autoregressive LLMs will never reach AGI because they lack world models, common sense, and physical embodiment. Promotes world-model-based architectures like V-JEPA."
Turing Award laureate (2018), pioneer of Convolutional Neural Networks (CNNs), and VP & Chief AI Scientist at Meta. LeCun advocates for open-source foundation models (Llama series) and critiques pure autoregressive LLMs as insufficient for true human-level intelligence.